Business analytics. Big data. Dashboards. Data mining. Management’s new favourite buzzwords. With the help of the press, analytics has become mainstream: the movie Moneyball is a good example. The analysis of data generated from CRM, ERP, social media, and POS systems provides significant insights. Have you ever wondered if you took the time to learn to use analytics, how it could help improve your business?
Recently, Harvard Business Review (Oct. 2012) highlighted the increasing importance of analytics, big data, and data scientists in many modern businesses. Despite entering the social consciousness, very few companies have embraced these concepts, and in my experience, fewer still do it well.
To help you get started, here is a quick how-to guide, based off “Making Advanced Analytics Work for You” by Dominic Barton and David Court.
Research
Start simple. Identify what information you have and what information you trust. Avoid trying to create data where there is none or cleaning polluted data through new governance/process initiatives. This is challenging and time consuming. Note: You may need to collect data from outside your systems. And you can always systemize important/useful data at a later date.
Formulate
Too many expect that by implementing analytics, the resulting data will magically reveal all the world’s secrets—it doesn’t work that way. You need to identify and define the business problem you are trying to solve, ensuring that the result of your analysis is actionable. Create a hypothesis. It is better to come armed with a specific question and look for an answer.
Design
Make sure your analytics are aligned with how and when your organization makes decisions. Will the data be used in quarterly planning sessions, or will it provide a metric that is reviewed weekly/daily? Understand the difference between data as a leading indicator or a trailing indicator.
Test
Collect baseline/control data as you will need something to compare against. Make a change (that will test your hypothesis) and see what happens. Remember, you want results that are actionable. When analyzing your data, you should be able to answer, “Did that change make a difference?”.
Analyze
Analyze the resulting data. Does it support your hypothesis? Yes/No–either way you will have learned something.
Conclude
If your results support your hypothesis, the next set is determining how you embed the knowledge and/or analytic tools into the systems and processes that manage the business. If they don’t, determine if your hypothesis was false or whether there were uncontrolled forces/factors that affected the results.
Does this process sound familiar? It should. The steps above are loosely aligned with the Scientific Method (think Grade 8 Science class). So if you approach analytics like you would an experiment, you’re simply formulating a hypothesis and analyzing the resulting data to test it.
If you’re interested in learning more about analytics, I recommend a great book “Competing on Analytics: The New Science of Winning” by Tom Davenport. But if you want to increase your firm’s analytical capabilities right away, be sure to get in touch.
